|Publication number||US5670379 A|
|Application number||US 08/372,708|
|Publication date||Sep 23, 1997|
|Filing date||Jan 13, 1995|
|Priority date||May 19, 1992|
|Publication number||08372708, 372708, US 5670379 A, US 5670379A, US-A-5670379, US5670379 A, US5670379A|
|Inventors||Masahito Ito, Junkichi Miura, Yoshio Fujii, Hiroshi Satake, Mitsuo Ito, Fuminori Umesato, deceased|
|Original Assignee||Hitachi, Ltd., Hitachi Instruments Engineering Co., Ltd.|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (6), Non-Patent Citations (2), Referenced by (15), Classifications (25), Legal Events (4)|
|External Links: USPTO, USPTO Assignment, Espacenet|
Pmax =P(S0)+(dP/dS)S=S.sbsb.0 (Smax -S0)+k(d2 P/dS2)S=S.sbsb.0 (Smax -S0)2
Pmax =P(S0)+(dP/dS)S=S.sbsb.0 (Smax -S0)+k(d2 P/dS2)S=S.sbsb.0 (Smax -S0)2
This application is a continuation of application Ser. No. 08/061,130, filed 13 May 1993, now abandoned.
The present invention relates to a chromatograph system.
Automation in chromatograph systems has been long desired. There is disclosed in JP-A-1-250060 a chromatograph system which automatically determines optimum analysis conditions in short time. According to this system, a plurality of preset analysis conditions and analysis data (retention times, peak dispersions, detector response times, and the like) of a plurality of kinds of sample component compounds separated under various analysis conditions are stored in a memory in advance. In analyzing a new sample, upon input of a desired resolution rate of a particular compound of the new sample, one of the analysis conditions suitable for analyzing the particular compound at the desired rate is selected by reading the stored analysis data from the memory and processing it in a predetermined manner. In this system, the analysis data may be stored in the memory in the form of spreadsheet, allowing addition and/or correction of the data.
There is also disclosed in U.S. patent application Ser. No. 07/836,599 filed on Feb. 18, 1992, now abandoned, corresponding to JP-A-4-331369 a chromatograph system which automatically identifies component compounds of a sample. JP-A-4-331369 was laid-open in Japan before the priority period of this application. According to this system, peaks of component compounds to be analyzed are identified by using chromatogram data of a standard sample with known components, and the width of a time window of each peak is pre-set. The windows are used for identifying the components of an unknown sample based upon its chromatogram. With this system, it is not necessary for an operator to enter the retention time of each component. The method of separating and identifying an unknown sample in this way is called a time window method.
Setting time windows is carried out at each run. During each run, thirty to one hundred unknown samples are analyzed in a general case. At each run, a standard sample is analyzed to set time windows, and thereafter, all unknown samples are sequentially analyzed without any interruption.
There is also disclosed in U.S. patent application Ser. No. 07/763,203, now abandoned, corresponding to JP-A- 4-130271 a chromatograph system which automatically displays the degree of deterioration of the column, eluent, derivating reagent, and the like to alarm the operator. With this system, the deterioration is judged, for example, from whether a standard retention time of a particular peak of a standard sample at each run becomes shorter or longer than a predetermined time period.
It is an object of the present invention to provide a chromatograph system with advanced automation.
It is another object of the present invention to provide a chromatograph system capable of analyzing with high reliability.
In order to achieve at least one of the above objects of the present invention, the conditions for identifying the peaks of a chromatogram are updated by referring to the previous analysis results stored in a memory. According to one embodiment, in a chromatograph system using the time window method whereby the peak identifying conditions are set by using a standard sample at each run, i.e., for each batch of unknown samples, a regression line regarding retention times of predetermined peaks measured at each run is obtained, and the peak identifying conditions are time sequentially corrected by referring to the regression line. The regression line indicates a run-to-run variation in retention time of the predetermined peaks.
According to another aspect of the present invention, a replacement time of a component having a limited life time in the chromatograph system is notified in advance to an operator. A column which is one of the system components has an intimate relationship between its life time and column pressure (resistance). The pressure can be used therefore as an index for the life time management. Conventionally, the pressure is monitored always to judge that the life time of the column came to the end when the pressure reached a predetermined value. The present inventors have found that there is a fixed rule of a column pressure change with the number of analyzed samples, i.e., with the lapse time from the start of using the column. This rule is applied to a change in measured column pressures to estimate the future column pressure and hence the column life time.
FIG. 1 is a block diagram showing the structure of a chromatography system for analyzing glycated hemoglobin according to an embodiment of the present invention.
FIG. 2 is a diagram showing the distribution of differences between measured and calculated retention times of peak F.
FIG. 3 is a graph showing the relationship between the measured retention times of peak F and peak A1c.
FIG. 4 is a graph showing a change in column pressures used as the criterion of column life prediction.
FIG. 5 is a graph showing a baseline drift at the start-up of the system.
FIG. 6 is a graph showing a change in noise generation frequencies at the start-up of the system.
Embodiments of the present invention will be described with reference to the accompanying drawings.
FIG. 1 shows the structure of a liquid chromatograph system for analyzing glycated hemoglobin according to an embodiment of the present invention.
Referring to FIG. 1, a pump 40 alternately feeds an eluent (A) 48, eluent (B) 49, and eluent (C) 50 respectively for 1.9, 1.0 and 0.4 minutes at the cycle time of 3.3 minutes, for performing the stepwise (gradient) elution.
A sampler 43 having a movable needle sucks 5 μl of a hemoglobin (Hb) standard sample, dilutes it through hemolyzing, and sends it to a path leading to a column 41. The standard sample 52 as well as the eluent (A) 48 is supplied to the column 41 where the components of the standard sample are separated and identified by a visible photometer 42. The detected data or chromatogram is stored in a memory of a data analyzing unit (data processing unit) 45. Reference numeral 51 represents unknown samples. For the hardware configuration, a glycated hemoglobin analyzer Type L-9100 manufactured by HITACHI, LTD., may be used.
The column 41 is of the same type as disclosed in U.S. patent application Ser. No. 07/896,784 filed on Jun. 9, 1992, now U.S. Pat. No. 5,294,336, which is a continuation application of U.S. patent application Ser. No. 07/578,214 filed on Sep. 6, 1990, now abandoned, and incorporated herein by reference.
The time window method using the system shown in FIG. 1 has been disclosed in U.S. patent application Ser. No. 07/836,599 filed on Feb. 18, 1992, now abandoned, which is incorporated herein by reference. An example of time windows is shown in Table 1.
TABLE 1______________________________________ Retention +/- AllowableWindow Peak Time (min) Width (min)______________________________________1 A1a 0.30 0.152 A1b 0.30t + 0.06 0.203 F 0.65t - 0.11 0.154 1-A1c 1.00t - 0.20 0.155 A1c 1.60 0.406 A0 2.60 0.30______________________________________
As described above, a standard sample with known components is analyzed before starting each run. On one of the conditions that the retention time of peak A1a of the standard sample is not within the range from 0.15 to 0.45 minutes, that the retention time of peak A1c is not within the range from 1.20 to 2.00 minutes, and that the retention time of peak A0 is not within the range from 2.30 to 2.90 minutes, the analysis of unknown samples will not be performed but an alarm is issued to the operator. This technique is being disclosed in U.S. patent application Ser. No. 07/763,203 filed on Sep. 20, 1992, now abandoned, which is incorporated herein by reference. The retention times of peaks A1b, F, and 1-A1c are obtained using the retention time of peak A1c. For example, assuming that the retention time of peak A1c is 1.85 minutes, the retention time of peak A1b is 0.30*1.85+0.06=0.62 minutes. Similarly, the retention times of peaks F and 1-A1c are calculated to be 1.09 and 1.65 minutes, respectively. In this manner, six windows are set for the analysis of unknown samples. If there are a plurality of peaks in one window, the object peak is selected by the following rules:
(1) A peak having the longest retention time is selected as peak A1a in window No. 1.
(2) A peak having the longest retention time is selected as peak A1b in window No. 2.
(3) A peak having the largest area is selected as peak F in window No. 3.
(4) A peak having the largest area is selected as peak 1-A1c in window No. 4.
(5) A peak having the largest area is selected as peak A1c in window No. 5.
(6) A peak having the largest area is selected as peak A0 in window No. 6.
Of the function 0.65t-0.11 for calculating the retention time of peak F shown in the time window Table, the coefficient 0.65 of the first order term and the coefficient--0.11 of the zero order term are determined experimentally. Similarly, the coefficients of the functions for calculating the retention times of peaks A1b and 1-A1c are determined. These coefficients are not, therefore, perfectly reliable for the practical analysis where the system, column, and eluent change with each lot and there is some deterioration of the column or the like.
This embodiment discloses the method of correcting these coefficients, or the method of correcting time windows. Data analysis conditions (functions practising the time window method) under which the peaks are identified by the data analyzing unit 45, are set by, and stored in, an analysis condition determining unit 53. A calculation section of the analysis condition determining unit 53 calculates to correct the coefficients of the data analysis conditions (functions) used by the time window method.
Table 2 shows the data of past five standard samples 52 obtained until the analysis of a new run starts. The data, including measured retention times T1 of peak A1c, measured retention times T2 of peak F, calculated retention times T3 of peak F from the function (0.65t-0.11), and differences between T2 and T3, at respective time and date, is stored in a secular change storage unit 55.
TABLE 2______________________________________Run 1 2 3 4 5______________________________________Measured 8/10 8/9 8/9 8/7 8/5Time/Date 11:02 15:15 10:57 12:00 9:00Measured 1.80 1.50 1.60 1.70 1.40RetentionTime T1 ofA1cMeasured 1.09 0.85 0.94 1.00 0.84RetentionTime T2 of FCalculated 1.06 0.87 0.93 1.00 0.80RetentionTime T3 of FT2 -T3 +0.03 -0.02 +0.01 ±0.00 +0.04______________________________________
FIG. 2 is a diagram showing the distribution of differences between measured and calculated retention times T2 and T3. As seen from FIG. 2, T2 is slightly longer than T3.
In a first embodiment, for the correction of the coefficients, the analysis condition determining unit 53 automatically selects five sets of past data from the secular change storage unit 55 to obtain an average value of differences, or a shift amount, and corrects the zero order term coefficient of the function. In this example, the average value +0.22 of differences is added to the initial coefficient--0.11 to change it to a new coefficient--0.09 and obtain a new function 0.65t-0.99.
In a second embodiment, in calculating the average value of differences, the differences are weighted greater for the later measured results.
For example, the shift amount is obtained by the equation (1). In this example, from the calculation given by (2), the function is corrected to 0.65t-0.10. ##EQU1##
This weighting method gives prominence to the later measured results, and is therefore practical from the viewpoint of eliminating the past influence of change in the column and eluent. If the number of past measurements is increased to be more than 5, the average value is less affected by the old measured results. The average value thus obtained is indicated by a broken line (regression line) in FIG. 2.
The automatic correction of the data analysis condition described above may be displayed in the form of a message on a monitoring means such as a CRT 46.
Alternatively, Table 2 or the diagram of FIG. 2 may be displayed on CRT 46 or printed out to a printer 47 for the confirmation by an operator. If the data automatically selected is improper or insufficient, an operator enters an identification number to delete, change, or add data.
In a third embodiment, both the first and zero order term coefficients are corrected. To this end, a regression line R between the measured retention times of peaks F and A1c is obtained as indicated by a broken line, to set the T2 -axis intercept of the line R to the zero order term coefficient, and the slope of the line R to the first order term coefficient. The solid line in FIG. 3 is defined by experimentally obtained coefficients. In correcting the coefficients in this manner, the number of data sets is preferably more than 5 in order to avoid a radical change of the first order term coefficient. The regression line is obtained, for example, by the least squares method. The data to be used is sequentially updated to obtain the latest data.
If the obtained correction coefficients change radically from the initial coefficients, there may be a correction error. In such a case, an alarm is issued. As the criterion for issuing an alarm, the time when the first order coefficient changes 20% or more, or the time when the zero order coefficient changes 0.20 minutes or longer, may be set.
In the above description, the sub-peak F and peak A1c have been used by way of example. A1b and 1-A1c may also be used for the coefficient correction of the functions. In the case of 1-A1c, however, this peak may not be detected from the standard sample 52. In such a case, data of an unknown sample 51 is used for the coefficient correction, and the analysis condition determining unit 53 processes the data and sends the corrected coefficients to the data analysis unit 45 for the coefficient setting.
The allowable width of each time window may be corrected. For example, the allowable width +/-0.15 minutes of peak F may be changed to a different width such as +/-0.10 and +/-0.20 minutes in accordance with the retention time variation of measured peaks F of several hundreds unknown samples 51. If the allowable width is to be magnified, the magnified width is first used to test whether peak F can be identified properly.
It is also possible to check the functional difference between time windows before and after correction. Namely, one or more peaks to be identified is subjected to the time windows before and after correction. If the peak is identified with the former but is not identified with the latter, or in a reverse case, then an operator is notified of an alarm. In this comparison, retention time of the peak and the time windows are stored in an outer or inner memory device.
Returning back to Table 2, it can be understood that the measured retention time of peak A1c changes greatly although it falls within the allowable range of 1.60+/-0.40 minutes. It has been found from the studies by the present inventors that the retention time of peak A1c of glycated hemoglobin becomes shorter as the number of runs increases. In another embodiment accordingly, an average value of five to twenty sets of past data is calculated for peak A1c in the same manner described above, to update the time window table. The peaks A1a and A0 are calculated also in the same manner.
As described above, the reliability of analyzed results improves as the time window table is updated, while alleviating the burden on an operator. In order to further improve the reliability of analyzed results, it is preferable to narrow the time window width, i.e., the allowable width shown in Table 2. This can be achieved by the embodiment described above. A high reliability can be achieved by using time window tables before and after correction and comparing the analyzed results.
The analysis condition determining unit 53 of the chromatograph system of the above-described embodiments is provided with another function of automatically determining a maintenance condition and automatically correcting a system control condition.
First, the function of automatically determining a maintenance condition will be described, in which a life time of a system component is estimated and a replacement time is notified to an operator. For example, in the case of estimating a life time of a column, the measured column pressure is used as a life time index, and the life time of the column is estimated from a change of the column pressure with time by using various life time functions.
The column pressure tends to rapidly increase as the number of analyzed samples increases, as shown in FIG. 4. In one embodiment of the life time estimating method, the number S of samples when the column pressure reaches its upper limit Pmax is calculated by using a first order differential value dP/dS (bar/sample) and second order d2 P/dS2 (bar/sample2) at a certain number So of samples, as given by the equation (3): ##EQU2##
For example, assuming that Pmax is 100 bar, that the pressure P(So) is 70 bar when the number So of samples reaches 400, that the first order differential value (dP/dS)S=So is 0.2 (bar/sample), and that the second order differential value (d2 P/dS2)S=So is 0.002 (bar/sample2), the number S of samples when the column pressure reaches the upper limit pressure Pmax is 500 as given by:
During the period while the 400-th sample is analyzed, it can be estimated that the column pressure will reach its upper limit after 100 more samples. The message of the column replacement time may be displayed on CRT 46 to notify the operator. This method is also applicable to not only a column replacement but also to an in-line filter replacement or card column replacement. However, it is difficult to estimate the replacement time during the period where the first and second order differential values take a value near zero. These differential values will not have large errors.
An example of calculating the differential values are given by the equations (5): ##EQU3##
Three data points are required for the second order differentiation. The three data points are represented by (S1, P1), (S2, P2), and (S3, P3) in the order of the latest, later, and late points.
In estimating the life time, the first order differential value at S=S1 is approximated to (dP/dS)S =S1,2, and the second order differential value is approximated to (D2 P/dS2)S =S1,2,3.
A predetermined function is prepared and corrected at certain time points using past measured data. For example, consider a function P=Co+C1 exp(C2 S) where P is a pressure (bar), S is the number of samples, Co(bar), C1 (bar), and C2 (1/sample) are coefficients. The number of samples is reset when the column 41 is replaced by a new one, and starts being counted. The default values of the coefficients are C1 =10 (bar) and C2 =0.003 (1/sample). The initial value of Co is (initial pressure (bar)--C1). Prior to starting the data sampling, the pump is required to be idled and the pressure after 10 minutes from the operation start of the pump is recorded at each run. An example of the measured data is shown in Table 3, the data being stored in the secular change storage unit 55.
TABLE 3______________________________________Run 1 2 3 4 5 6 7______________________________________Date 8/1 8/2 8/2 8/3 8/4 8/6 8/7Pressure 50 51 53 53 55 56 58Number of 0 30 80 90 120 150 190analyzedsamples______________________________________
The coefficients of the function are corrected, for example, at the seventh run, by using the pressure data and the numbers of analyzed samples of the past seven runs, to match the past history of data change. In the example of Table 3, the number of samples when the pressure reaches its upper limit of 100 bar can be estimated to be 500. At the end of the seventh run, there is displayed a message that the number of samples which can still be analyzed is 310.
The analytical function has been used for the estimation in the above embodiment. Such a particular function is not necessary, but a function representing a typical pattern of representing a pressure increase may be used. An example of this function is a monotone increasing function y=f(x), where y represents the pressure P and x represents the number S of samples. Using this function, P is related to S by:
P=Co+C1 f(C2 S) (6)
where Co(bar) is a constant, C1 (bar) is a coefficient representing the degree of pressure change, and C2 (1/sample) is a coefficient representing a speed of pressure change per sample.
Discrete sets of values (xi, yi) are stored. The number of points xi is about 20. Each value xi is related to a corresponding value yi. The value between xi and x(i+1) is obtained by interpolation. In just the same manner as the analytical function method, the number of samples is counted after the column 41 is replaced, and the Co, C1, and C2 are corrected at each data measuring point by using the past pressure data in order to match the past history of data change.
The function f(x) is automatically corrected in accordance with the measured data y. The measured discrete sets of values (xi, yi) are replaced by new sets of values including interpolated values which are obtained by using the matched Co, C1, and C2. The measured values (xi, yi) may be weighted by 10% and added to the already obtained values (xi, yi) representing the pressure change pattern so that the pattern is always refreshed to reflect the actual pattern.
The following items may also be estimated.
(1) Retention time: With the increasing or decreasing tendency of the retention time, the time when the retention time exceeds the allowable range can be estimated by the differential coefficient method to request column replacement.
(2) Peak width: An increase of the width (peak area/peak height) of peak A1c of the standard sample 52 can be estimated by the differential coefficient method to request column replacement.
(3) Lamp intensity: With the monotonous increase of lamp intensity, the limit time of the lamp can be estimated by the differential coefficient method to request lamp replacement.
(4) Detector noise/drift value: The time when these values become stable can be estimated by the differential coefficient method to notify the sampling start time.
When the chromatograph system starts its operation, a lamp in the detector turns on. Until the temperature of the lamp becomes stable, the baseline of the detector drifts (refer to FIG. 5) and much noise is generated (refer to FIG. 6).
With a conventional technique, the analysis of standard samples starts when both the baseline voltage and the noise occurrence frequency lower to predetermined threshold values or less.
In the present embodiment, the times when the baseline voltage and the noise occurrence frequency reach the threshold values can be estimated by the differential coefficient method or the like. In the example shown in FIGS. 5 and 6, times TT1 and TT2 when the baseline voltage and noise occurrence frequency become stable are estimated at time tF, and the later one of the times TT1 and TT2 is selected as an analysis start enabled time which is notified to the operator. With such a function, a case where the system does not become stable can be also estimated.
The estimation of the analysis start enabled time may be made using either the baseline drift or the noise occurrence frequency as an index.
The above-mentioned estimations are done by the analysis condition determining unit 53. The result thus obtained is delivered to a maintenance unit 54 and a message is issued.
The life time management index may reach the limit value not from its life time itself but from other causes. Such causes may be identified from the secular change data. An example of such identification will be described. Consider for example the chromatograph system wherein the lamp intensity is measured as the deterioration index of the lamp in the detector 42. The lamp intensity decreases monotonously generally in the unit of 100 hours. In this system, an amplifier is used which sets the intensity of a newly replaced lamp to 100 to 200 mV. The lamp intensity is measured at a fixed time after 10 minutes from the start of the pump operation. When the lamp intensity lowers to a predetermined value 50 mV or less, an alarm is issued to the operator to request lamp replacement.
In this embodiment, however, an alarm is not necessarily issued even if the lamp intensity lowers to 50 mV or less. Instead, the maintenance unit 54 first refers to the secular change storage unit 55 to check the secular change of the lamp intensity. If the present value at the predetermined value of 50 mV or less changes abruptly by 20% or more than the value measured at the preceding point, then it is not judged that the lamp life came to the end, but a message of a possibility of a contaminated flow cell is displayed on CRT 46. If the present value at the predetermined value of 50 mV or less does not lower by 20% or more, it is judged that the lamp intensity has lowered gradually to the predetermined value, and so a message of a replacement by a new lamp is displayed.
There is an alternative method of more precisely judging whether the lamp has been deteriorated by its own life time or other causes. Namely, the decrease of the lamp intensity is estimated by the differential coefficient method. If the lamp intensity becomes lower than the estimated value by -10% or more, it is judged that there is another cause except the simple lamp deterioration. Also in the case where the lamp intensity becomes higher than the estimated value by +10% or more, there is a possibility of an abnormal state of other elements.
An abrupt rise of the column pressure can also be checked in the similar manner as above. The retention time, peak width, and injection amount by the sampler 43 using a standard sample having same component compounds, may also be checked in the similar manner.
Lastly, the function of automatically correcting the control conditions will be described. With this function which changes the control conditions slightly, the normal analysis can be continued for a longer time. For example, the case where the retention time of peak A1c of the standard sample becomes gradually short, may be reasoned from the deterioration of the column 41. In such a case, if the pump flow is lowered by 10% when the retention time of peak A1c becomes 1.30 minutes or shorter (this is a second time window), the retention time of peak A1c can be elongated by about 10%. As an alternative of this, the switching time between the eluent (A) 48 and eluent (B) 49 may be delayed by 0.2 minutes.
As a more improved alternative, the control conditions such as flow rate, switching time, column temperature, or in some case, mixture of eluents may be changed slightly so as to regulate the retention time of peak A1c to near 1.7 minutes. In this case, the following two points should be taken into consideration.
(1) Parameters used for peak identification are corrected time sequentially in the manner described above.
(2) A rule is introduced wherein each control condition is slightly changed while suppressing an abrupt change of peak identification parameters, and the condition which provides the most effective results is selected. The chromatograph system itself may be made to learn such a rule.
Specifically, the results obtained in response to a change of the control condition are stored in the secular change storage unit 55, and referred to when the next control condition change is performed. By storing the secular change of the control condition, it is possible to use it for estimation at the next control condition change. The operation of the control condition change is conducted by the analysis condition determining unit 53 and executed by the control unit 44.
Using the secular change data, the embodiments of the present invention provide the following advantageous effects.
(1) The coefficients of a peak identifying function used for the data analysis can be corrected so that a difference between columns, a lot variation of eluents, and a secular change of measured values caused by system components deterioration can be flexibly dealt with.
(2) The life time management index of each system component can be estimated for system maintenance so that a replacement time of each system component can be notified to an operator. Furthermore, a cause other than the expiration of life time can be discriminated when the life time management index fluctuates greatly.
(3) By slightly changing the control condition, it is possible to elongate the replacement time of a system component having a limited life time.
Determination and correction of various conditions necessary for the proper analysis improve the fidelity, precision, and reliability of the analyzed data.
According to one aspect of the present invention, the data analysis condition is automatically corrected in accordance with the secular change of measured values, improving the data analysis precision and ensuring the proper analysis.
According to another aspect of the present invention, the life time is estimated more correctly and the proper component replacement time can be notified. When the life time index takes a predetermined value, it is possible to determine whether it has occurred from the life time itself or from another cause, further improving the system reliability.
According to a further aspect of the present invention, by changing the control condition, it is possible to elongate the component replacement time.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US3898837 *||Sep 17, 1973||Aug 12, 1975||Dietrich A H Boege||Method of and device for the identification and reduction of peaks in chromatograms|
|US4197576 *||Aug 4, 1977||Apr 8, 1980||Juan Martin Sanchez||Adaptive-predictive control system|
|US4674323 *||Jul 31, 1985||Jun 23, 1987||The Dow Chemical Company||Self-diagnostic gel permeation/size exclusion chromatograph|
|US4802981 *||Nov 4, 1987||Feb 7, 1989||Oros Systems Limited||Automatic chromatography apparatus|
|JPH04130271A *||Title not available|
|JPH04331369A *||Title not available|
|1||*||Dolan, John W. Troubleshooting LC Systems, The Humana Press Inc, Clifton NJ 1989 p. 18, 22 24, 74.|
|2||Dolan, John W. Troubleshooting LC Systems, The Humana Press Inc, Clifton NJ 1989 p. 18, 22-24, 74.|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US5827946 *||Apr 30, 1997||Oct 27, 1998||Hewlett-Packard Company||Method for sample identification using a locked retention time database|
|US5905192 *||Jul 23, 1997||May 18, 1999||Hewlett-Packard Company||Method for identification of chromatographic peaks|
|US5915269 *||Apr 14, 1998||Jun 22, 1999||The Perkin-Elmer Corporation||Method and apparatus to compensate for gas chromatograph column permeability|
|US5939612 *||Apr 28, 1998||Aug 17, 1999||Hewlett-Packard Company||Retention time-locked spectral database for target analyte analysis|
|US5987959 *||Mar 6, 1998||Nov 23, 1999||Hewlett-Packard Company||Automated retention time locking|
|US6036747 *||Jul 24, 1998||Mar 14, 2000||Hewlett-Packard Company||Column specific parameters for retention time locking in chromatography|
|US6210465 *||Mar 17, 1999||Apr 3, 2001||Agilent Technologies, Inc.||Method for identification of components within a known sample|
|US6315901 *||Mar 17, 2000||Nov 13, 2001||Shimadzu Corporation||Liquid chromatography|
|US6438499 *||Sep 10, 1999||Aug 20, 2002||Tosoh Corporation||Chromatogram analyzer|
|US6494078 *||Jun 25, 2001||Dec 17, 2002||Agilent Technologies, Inc.||Retention-time locked comprehensive multidimensional gas chromatography|
|US6736975||Dec 3, 2001||May 18, 2004||Scynexis, Inc.||Method of separating compound(s) from mixture(s)|
|US20040164024 *||Mar 1, 2004||Aug 26, 2004||Thierry Gandelheid||Method of separating compound(s) from mixture(s)|
|US20130228677 *||Feb 6, 2013||Sep 5, 2013||Hitachi High-Technologies Corporation||Mass Spectrometric System|
|EP1342202A1 *||Dec 5, 2001||Sep 10, 2003||Waters Investments Limited||Automated test protocol|
|WO2002047009A1 *||Dec 5, 2001||Jun 13, 2002||Waters Investments Ltd||Automated test protocol|
|U.S. Classification||436/161, 436/66, 73/23.22, 210/656, 73/61.52, 210/198.2, 422/89, 95/82, 422/70, 96/103, 73/23.36|
|International Classification||G01N30/86, G01N30/62, G01N30/02, G01N30/82, G01N30/88|
|Cooperative Classification||G01N2030/027, G01N2030/8804, G01N2030/626, G01N30/8631, G01N2030/625, G01N30/82, G01N30/88|
|European Classification||G01N30/82, G01N30/88|
|Nov 10, 1998||CC||Certificate of correction|
|Feb 27, 2001||FPAY||Fee payment|
Year of fee payment: 4
|Feb 25, 2005||FPAY||Fee payment|
Year of fee payment: 8
|Feb 26, 2009||FPAY||Fee payment|
Year of fee payment: 12